Research On The Rice Yield Prediction Model In Yunnan Province Based On Data Mining

With the rapid growth of global population,food security has gradually become one of the core issues in every country.Scientific prediction of food production,as an important method to ensure national food security,has become a hot research direction in the field of agricultural sciences in various countries.As one of the world’s three major food crops,rice is the main source of food for most of the world’s population.Therefore,accurate prediction of rice production is of great significance for ensuring food supply and national food security.This study is based on data mining techniques and web development technologies to construct a rice production prediction model and a rice production prediction system for Yunnan Province.The specific research content and results are as follows:(1)Using Yunnan Statistical Yearbook as the data source,association rule analysis and grey relational analysis are used to analyze the correlation between 7 influencing factors,including total reservoir capacity,nitrogen fertilizer usage,phosphorus fertilizer usage,potassium fertilizer usage,compound fertilizer usage,pesticide usage,and rice planting area,and rice production.Strongly correlated factors are selected as sample features for the rice production prediction model.The research results show that total reservoir capacity,nitrogen fertilizer usage,phosphorus fertilizer usage,pesticide usage,and rice planting area have strong correlations with rice production and are more suitable as sample features for constructing the rice production prediction model.(2)Four data prediction techniques,including multiple linear regression,CART tree,random forest,and BP neural network,are used for modeling and predicting rice production in 16 regions of Yunnan Province.Based on the comparison of model performance,the optimal models from all regions are integrated into an integrated model for predicting rice production in Yunnan Province.Verification results show that the data mining-based integrated model for predicting rice production in Yunnan Province has better prediction performance than multiple linear regression,CART tree,random forest,and BP neural network models,with a Mean Absolute Error(MAE)of 71,100 tons,Mean Squared Error(MSE)of 6,256,100 tons^2,Root Mean Squared Error(RMSE)of 79,100 tons,and Mean Absolute Percentage Error(MAPE)of 1.6%.The model has high applicability and generalization performance,and can accurately predict rice production in16 regions and the whole province of Yunnan.(3)HTML+CSS+Java Script+AJAX technologies are used to build the rice production prediction system for Yunnan Province.HTML+CSS+Java Script are used for front-end page development,and AJAX is used for asynchronous refresh and data requests,making the model more efficient and providing a better user experience.Python is used for back-end data processing and modeling,ensuring the efficiency and accuracy of model calculation.My SQL database is used to provide database services for the system,ensuring data storage integrity and security.Through system construction,users can conveniently and quickly use the rice production prediction model for Yunnan Province,thus providing better guidance for the development and macro-regulation of the rice industry in Yunnan Province.